phenomena cohered in 1880. The remaining five have already been discussed. The conclusion is that in at least twenty-five of the States and Territories the density and the birth-rate per thousand women between the ages of fifteen and forty-nine are opposed.

The comparison of farm values per acre of improved land with the birth-rates for 1890 shows apparently an exact coincidence with the results of the similar comparison for 1880. In sixteen States and Territories the phenomena cohere; they oppose each other in thirty (Tables VIII and VI). In fourteen of the forty-six States and Territories, however, the phenomena have changed their relative positions, but in such a way as to make the total result the same: in seven States and Territories (New Mexico, North Carolina, California, Maine, New Hampshire, Vermont, and Washington), the phenomena cohered in 1880 and are opposed in 1890; in seven States and Territories (Illinois, Indiana, Iowa, Montana, Nevada, Oregon, and Wyoming), they were opposed in 1880, and cohere in 1890. In six of these States and Territories (Illinois, Iowa, Montana, Oregon, Washington, and Wyoming), the variation results from the change in the birth-rate which has already been explained in comparing the statistics of birth-rate and density for 1880 and 1890. The variation in the remaining eight States and Territories is caused by the change in farm values: in four States and Territories (Indiana, Nevada, New Mexico, and North Carolina), the farm values per acre of improved land were above the average for the United States in 1880, and are below it in 1890; and in four States (California, Maine, New Hampshire, and Vermont), the opposite change has taken place. If the four States (Montana, Oregon, Washington, and Wyoming) whose birth-rates, if computed on the same basis as those of 1880, would probably have had a different relation to the average rate for the United States, be considered to vary in relation to farm values as they did in 1880, it will be seen that thirty-six of the forty-six States and Territories show the

same relations of the phenomena of birth-rate and farm values in 1890 that they did in 1880. This coincidence is but another proof that there is some dependence of the one upon the other.

A comparison of the birth-rate with the values of manufactured products per capita in 1890 (Table VIII.) shows twelve States and Territories in which the phenomena cohere and thirty-four in which they oppose each other. If the four States (Montana, Oregon, Washington, and Wyoming), whose birth-rate is estimated as above the average for the United States when computed on the basis of the number of women between the ages of fifteen and forty-nine, be so counted, the phenomena of the birth-rate and the values of manufactured products per capita will oppose each other in thirty-eight of the forty-six States and Territories. The coherences and the oppositions for 1890 are almost identical with those of 1880; in five States only (Illinois, Indiana, Iowa, Maine, and the District of Columbia) is there a difference, if Montana, Oregon, Washington, and Wyoming be omitted. In Illinois and Iowa, the relative position of the birth-rate has changed; in Indiana and the District of Columbia, the value of manufactures per capita has in 1890 risen above the average for the United States; in Maine it has fallen below the average; therefore, in forty-one of the forty-six States and Territories the relative conditions of manufactures and the birth-rate are the same in 1890 as they were in 1880.

If the average rates for the United States in 1880 and in 1890 be compared, the results obtained from the preceding detailed comparisons are confirmed. The birth-rate has diminished from 30.95 per thousand of population to 26.68.* The value of agricultural products per acre of improved land has also decreased: in 1880 it was $7.77; in 1890, $6.88. The density per square mile of area of settlement has increased from 31.96 to 32.16, and the density per square mile

* Billings, "The Diminishing Birth-rate in the United States."-The Forum,

of total land surface, from 17.29 to 21.31. And, finally, the value of manufactured products has risen from $106.50 per capita to $149.63.

In his study of the conditions. of mortgage indebtedness in the United States, Mr. Holmes has shown that the mortgage debt, in general, increases with expanding prosperity. We should therefore expect it to show coherence with the density, the value of manufactured products, and to some extent with the values of agricultural products per acre of improved land, and opposition to the birth-rate. A comparison of the statistics proves the truth of this assumption. The phenomena of the birth-rate and the mortgage debt cohere in sixteen States and Territories (Table VIII), and are opposed in twenty-nine. Of the sixteen in which the phenomena cohere, three of the four mentioned above (Montana, Oregon, and Wyoming) must be put among those in which the phenomena oppose each other, thus making thirtysix in this class. The mortgage indebtedness is above the average for the United States in fifteen States (California, Colorado, Connecticut, the District of Columbia, Iowa, Massachusetts, New Jersey, New York, Pennsylvania, Rhode Island, Washington, Illinois, Kansas, Minnesota, and Nebraska). Nine of these have a value of manufactured products per capita above the average for the United States. The others are principally western farming States (Colorado, Iowa, Washington, Kansas, Minnesota, Nebraska).

In order that the relative rise or fall of the various factors chosen for comparison may be more clearly seen, the percentages of variation for 1880 have been calculated for the five great divisions given in the census reports:-the North Atlantic, the South Atlantic, the North Central, the South Central, and the Western divisions. The results are shown in Table IX. and Chart III. In every division the death-rate from nervous diseases coheres with the value of the manufactured products per capita, and both oppose the birth-rate. The North Atlantic division, which has by far the greatest amount

of manufacturing, has much the highest death-rate from nervous diseases and the lowest birth-rate. This division has also the greatest density and the highest intensity of agricultural cultivation, as represented by the value of agricultural products per acre of improved land. That it is not mainly an agricultural region is shown by the low rate of agricultural values per capita.

TABLE IX.

BIRTH-RATES AND FACTORS OF ECONOMIC CONDITION BY GRAND DIVISIONS, 1880.

With one exception only (the North Central division), the density per square mile of area of settlement coheres with the death-rate from nervous diseases and the values of the manufactured products, and opposes the birth-rate; and with one exception (the South Central division, which

Value of manufactured

products per capita.